Academic Paper: https://www.pnas.org/content/117/15/8398?etcc_med=newsletter&etcc_cmp=nl_algoethik_17907&etcc_plc=aufmacher&etcc_grp=

Even if this paper stated that the best predictions were not very accurate and were only slightly better than those from a simple benchmark model, researchers will continue to try to predict the future of humans. It is scary in a way because if a model could predict accurately or accurately enough what will happen to a family would fall into the hand of bad people, we would end up with a world which will be pretty similar to the life in the movie “Gattaca”. The only difference would be that in the movie, they could predict the evolution of a child by using a DNA test, and here it would be by using the data on his own family.

One of the obstacles to lifting in order to improve the prediction is already stated at the end of the article: “Ideally, these assessments would be carried out with government administrative data used in policy settings because the properties of these data likely differ from the properties of the Fragile Families data, but legal and privacy issues make it difficult for researchers to access many types of administrative data”. If researchers may not have access to some data that could help refine their predictive models, they may use other data set to achieve the same goals. We already know that the amount of data collected increases exponentially (90% of the world’s data has been created in the last two years alone) and will continue to do so. Alex Pentland, who also participates in this research, with his students, was “able to specify regular temporal and spatial patterns of location, activity, and communication­use patterns, which together enabled predictions of up to 90 percent accuracy regarding where someone was likely to be and what that person was likely to be doing within the hour, as well as highly accurate predictions about an individual’s colleagues, casual friends, and close relationships.” (the age of surveillance, Soshana Zuboff). So, personally, I think it is legitimate to ask ourselves: How long will it takes before researchers will have enough data and enough computational power to predict our life before we had the time to live it? How long will it take before our life becomes just a list of checkboxes, milestones that have been predicted? Would life have meaning if we knew what will happen next? What will happen if politicians implement an accurate model to give us benefits and restrictions based on this model?  Those questions may look like pure science fiction right now, but in the same way, there are laws to prevent genetically manipulations on humans, we should already set boundaries about what can be done with data and what cannot.